Closed Barley194 closed 9 months ago
There have been a lot of optimizations to recordings in 0.13 (currently in beta) so in general this should be resolved but may be indicative of slow segment moving times. I'd suggest reading https://deploy-preview-6262--frigate-docs.netlify.app/troubleshooting/recordings
Moving times are fast:
2023-10-26 04:57:11.733593418 [2023-10-26 04:57:11] frigate.record WARNING : Unable to keep up with recording segments in cache for reolink410WFirstFrigate. Keeping the 5 most recent segments out of 6 and discarding the rest... 2023-10-26 04:57:12.249325063 [2023-10-26 04:57:12] frigate.record DEBUG : Copied /media/frigate/recordings/2023-10-26/04/reolink410WFirstFrigate/56.20.mp4 in 0.44483208656311035 seconds. 2023-10-26 04:57:12.565578536 [2023-10-26 04:57:12] frigate.record DEBUG : Copied /media/frigate/recordings/2023-10-26/04/reolink410WFirstFrigate/56.29.mp4 in 0.2894909381866455 seconds. 2023-10-26 04:57:12.781152934 [2023-10-26 04:57:12] frigate.record DEBUG : Copied /media/frigate/recordings/2023-10-26/04/reolink410WFirstFrigate/56.39.mp4 in 0.2014169692993164 seconds. 2023-10-26 04:57:12.996955421 [2023-10-26 04:57:12] frigate.record DEBUG : Copied /media/frigate/recordings/2023-10-26/04/reolink410WFirstFrigate/56.49.mp4 in 0.19872617721557617 seconds. 2023-10-26 04:57:13.425259357 [2023-10-26 04:57:13] frigate.record DEBUG : Copied /media/frigate/recordings/2023-10-26/04/reolink410WFirstFrigate/56.58.mp4 in 0.2891509532928467 seconds. 2023-10-26 04:57:22.070929600 [2023-10-26 04:57:22] frigate.record DEBUG : Copied /media/frigate/recordings/2023-10-26/04/reolink410WFirstFrigate/57.10.mp4 in 0.19686198234558105 seconds. 2023-10-26 04:57:32.137783268 [2023-10-26 04:57:32] frigate.record DEBUG : Copied /media/frigate/recordings/2023-10-26/04/reolink410WFirstFrigate/57.19.mp4 in 0.2512190341949463 seconds. 2023-10-26 04:57:42.077874152 [2023-10-26 04:57:42] frigate.record DEBUG : Copied /media/frigate/recordings/2023-10-26/04/reolink410WFirstFrigate/57.29.mp4 in 0.19926190376281738 seconds. 2023-10-26 04:57:52.093959697 [2023-10-26 04:57:52] frigate.record DEBUG : Copied /media/frigate/recordings/2023-10-26/04/reolink410WFirstFrigate/57.38.mp4 in 0.18979191780090332 seconds. 2023-10-26 04:58:02.123380471 [2023-10-26 04:58:02] frigate.record DEBUG : Copied /media/frigate/recordings/2023-10-26/04/reolink410WFirstFrigate/57.48.mp4 in 0.21117496490478516 seconds. 2023-10-26 04:58:12.174261173 [2023-10-26 04:58:12] frigate.record DEBUG : Copied /media/frigate/recordings/2023-10-26/04/reolink410WFirstFrigate/58.00.mp4 in 0.27132701873779297 seconds. 2023-10-26 04:58:22.068202075 [2023-10-26 04:58:22] frigate.record DEBUG : Copied /media/frigate/recordings/2023-10-26/04/reolink410WFirstFrigate/58.09.mp4 in 0.17260289192199707 seconds. 2023-10-26 04:59:21.733324824 [2023-10-26 04:59:21] frigate.record WARNING : Unable to keep up with recording segments in cache for reolink410WFirstFrigate. Keeping the 5 most recent segments out of 6 and discarding the rest... 2023-10-26 04:59:31.747901673 [2023-10-26 04:59:31] frigate.record WARNING : Unable to keep up with recording segments in cache for reolink410WFirstFrigate. Keeping the 5 most recent segments out of 6 and discarding the rest...
And CPU usage is not high.
Gotcha, then yeah this is something that'll require running 0.13 to fix
This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.
Describe the problem you are having
After moving frigate installation from proxmox vm (ubuntu server + docker) to proxmox lxc (privileged, as I couldn't make the coral usb pass-through work on unprivileged one) + docker in order to reduce the coral intereference speed from ~60ms to ~15ms
or
after started using go2rtc restream for recording and detection
(I'm not sure which of those 2 actions have started it)
I get the:
Unable to keep up with recording segments in cache for reolink410WFirstFrigate. Keeping the 5 most recent segments out of 6 and discarding the rest...
spam all the time.
I run this setup for a few weeks now and this spam is all day and night long everyday.
It seams like everything what should, is recorded. (I don't watch every single recording but the ones I watched didn't looked cropped at the end for e.q. - despite "post_capture: 60" the recordings ends in less than 10s after the motion is no longer visible, but maybe that's how it works?) Sometimes something is not detected or detected but not recorded (not sure) but I don't know if this is related.
My storage for media is a network samba drive which is located on the same proxmox device but inside a VM. At first I thought that's the issue so I pointed the save location of recordings to a local drive on the LXC, but the logs kept spamming. So this is not it.
I have only 1 camera for now, it's Reolink 410W and is connected via WiFi. My host for frigate is connected via ethernet. Camera sometimes loses wifi signal, but this is shown as ffmpeg error in frigate logs - and that's OK.
I feel like I tried everything, searched the web, red through bunch of past issues here on github and nothing has helped me.
PS: As I don't see a big issue caused by this log I could I just ignore that. But I'm afraid that there is something wrong and thus I'd love to fix it + my logs are cluttered now.
Frigate LXC resources:
shm and cache:
I'm willing to provide any necessary information which I did not include here.
Version
0.12.1-367d724
Frigate config file
Relevant log output
FFprobe output from your camera
Frigate stats
Operating system
Proxmox
Install method
Docker Compose
Coral version
USB
Network connection
Mixed
Camera make and model
Reolink 410W
Any other information that may be helpful
No response